The Red River Valley Flood of 1997: A Call for Worst-Case

Scenario Approaches to Flood Risk Management

Paul E. TodhunterDepartment of GeographyUniversity of North Dakota

Grand Forks, ND, 58202-9020USA

Frequency-Magnitude-Damage Relationships

Extreme Flood Hydroclimatology

• Unique combination of flood-producing processes

• Not sampled in historical record• Flood Flow Frequency Analysis (FFFA)

inadequate to estimate Qpk100

Global Change

• Global climate change• Global environmental

change• Cultural change• Assumptions of FFFA not

valid

Clark (2005): Worst-case Thinking

• Preemptive resilience as a national disaster mitigation strategy

• Probabilism• Possibilism

1997 Red River Valley FloodGrand Forks, ND – East Grand Forks, MN USA

Source: USGS

Constant Flood Factors

Flood Forecast-Response System (FFRS)

(Krzysztofowicz, 1983)

1) Data collection

2) Flood forecasting

3) Forecast dissemination

4) Decision-making

5) Action implementation

FFRS – Step 1Data Collection

• United States Geological Survey• Real-time river stage observations• Telemetry input to NCRFC• Required NWSRFS input data

Stream Gage

FFRS – Step 2Flood Forecasting

• National Weather Service River Forecast Centers

• NWS River Forecasting System modeling system

• Numerical Outlook – climatological forecast• Operational Forecasts – NWSRFS for 34

Red River Basin forecast points

Variable Flood Factors

• Long lead time Fall soil

moisture Seasonally

frozen soils Snow water

equivalent of snowpack

• Short lead time Spring thaw Spring rain-on-

snow events River ice

1 6 11 16 21 26 31

April 1997

0

10

20

30

40

50

60R

ive

r S

tag

e (

ft)

NCRFC

RR-Grand Forks

CrestApril 22

02:00

Outlookmode

Forecastmode

“Stair-Stepping”

1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000

W ater Year

0

50

100

150

200

250

300

Sn

ow

fall

(cm

)GFU

1892-1997

Transect of snow depth and snow density in a prairie snowpackPomeroy et al. (1993)

1-M

ar

8-M

ar

15

-Mar

22

-Mar

29

-Mar

5-A

pr

12

-Ap

r

19

-Ap

r

26

-Ap

r

0

100

200

300

400

M

DD

(C

°)W ahpetonFargoGrand ForksPembina

Bliz

zard

1979 Flood

Historical Extrapolation

18 April 1997 Flood Statement, 2008 LST“This situation is unlike any flooding conditions ever experienced in eastern ND and NW MN. The NWS is working very hard with local, state, and federal agencies to give the public the most accurate information possible.”

USA Today, 14 June 1997, NCRFC Head

“... We really didn’t have an indication we were going to get 54 feet until all the pieces fell into place and we put the puzzle together.”

FFRS – Step 3Forecast Dissemination

• National Weather Service Eastern North Dakota Forecast Office

• Staff Hydrologist• Flood Statements addressed

hydrological uncertainty to an educated audience

FFRS – Step 4Decision-Making

• Local governments and agencies• City Emergency Manager• City Engineer• Mayor• Police, Firefighters, Sheriff, etc.

1997 “The Perfect Snowmelt Flood”

• 1979 Flood: 49 ft• Changes since 1979• Anchoring of perception

• Midplaced concreteness• Unfamiliar with NWS

products / hydrology• Narrow flood-defense strategy

NWSRFC

Forecast Point

2nd Numerical Outlook (feet)

14 March 1997

1997 Flood Peak

(feet)

Red River – Wahpeton 18.5 19.42

Red River – Fargo 37.5 39.72

Red River – Halstad 39.5 40.74

Red River – Grand Forks 49.0 54.35

Red River – Drayton 44.0 45.60

Red River – Pembina 54.0 54.90

Wild Rice River – Abercrombie 24.5 26.59

Sheyenne River – Lisbon 18.5 19.29

Sheyenne River – Kindred 21.0 22.33

Sheyenne River – West Fargo 22.5 23.30

Sheyenne River – Harwood 894.0 892.0

NWSRFC

Forecast Point

2nd Numerical Outlook (feet)

14 March 1997

1997 Flood Peak

(feet)

Maple River – Mapleton 14.50 15.50

Pembina River – Walhalla na 16.20

Pembina River – Neche 22.0 24.51

Buffalo River – Hawley 10.0 10.77

Wild Rice River – Twin Valley 13.5 15.90

Wild Rice River – Hendrum 31.5 33.85

Marsh River – Shelly na 25.70

Red Lake River – Crookston 27.0 28.40

Snake River – Alvarado 10.0 10.60

Two Rivers River – Hallock 810.0 810.7

-101 -100 -99 -98 -97 -96 -95 -94

45

46

47

48

49

50

Flood-of-record by 7 April

Flood-of-record by 15 AprilEventual flood-of-recordNCRFC Forecast Point

NWS NCRFC Forecast Points

25

-Ma

r2

6-M

ar

27

-Ma

r2

8-M

ar

29

-Ma

r3

0-M

ar

31

-Ma

r1-

Ap

r2-

Ap

r3

-Ap

r4-

Ap

r5

-Ap

r6

-Ap

r7

-Apr

8-A

pr

9-A

pr1

0-A

pr1

1-A

pr1

2-A

pr1

3-A

pr1

4-A

pr1

5-A

pr1

6-A

pr

0

25

50

75

100

125

150S

an

db

ag

s (

x1

00

0)

Bliz

zard

Ha

nn

ah

Dik

es F

ail

Emergency Sandbagging Operations

FFRS – Step 5Action Implementation

• Local workers• Federal workers – National Guard,

Army Corps of Engineers, Volunteers

EvacuationSites

Emergency OperationsCenter

10 km1

2

3 3 2 1← . .

.

.

. .

0 10 20 30 40 50 60 70 80 90 100 110 120

Julian D ay

0

10

20

30

40

50

60

Riv

er

Sta

ge

(ft

)0.90, 0.70, 0.50, 0.30, 0.10 Probabilities

Observed River Stage - 2006

Flood Stage

Outlookmode

Forecastmode

Advanced Hydrologic Prediction System

Caveant admonitus (Let the forewarned beware)

Walter M. Kollmorgen (1953) Settlement Control Beats Flood Control. Economic Geography, 29: 208-215.

Gilbert White(1911-2006)